Solar Vacuum-ultraviolet Radiometry with SUMER

Transcription

1 10 Solar Vacuum-ultraviolet Radiometry with SUMER KLAUS WILHELM, UDO SCHÜHLE, WERNER CURDT, INGOLF E. DAMMASCH Max-Planck-Institut für Aeronomie, Katlenburg-Lindau, Germany JÖRG HOLLANDT Physikalisch-Technische Bundesanstalt, Berlin, Germany PHILIPPE LEMAIRE Institut d Astrophysique Spatiale, Unité Mixte, CNRS-Université Paris XI, Orsay, France MARTIN C.E. HUBER ESA, Space Science Department, Noordwijk, The Netherlands currently at: International Space Science Institute, Bern, Switzerland Since the beginning of 1996, the space-based telescope and spectrograph SUMER (Solar Ultraviolet Measurements of Emitted Radiation) on the SOlar and Heliospheric Observatory (SOHO) of ESA and NASA has obtained spectra of many features of the quiescent and active Sun with high spectral and spatial resolution. In addition, irradiance and radiance measurements of line and continuum emission have been performed in the wavelength range 46.5 nm to nm. The instrument was radiometrically calibrated against the Berlin Electron Storage ring for SYnchrotron radiation (BESSY I), a primary source standard, with the help of a transfer source standard based on a hollow-cathode discharge lamp. A thorough cleanliness programme, specifically aimed at chemical contamination control, resulted in an excellent radiometric stability of the normal-incidence optical system as well as of the detectors. This has been verified under operational conditions by various techniques employed during the SOHO mission, such as line-ratio studies, observations of stars, and comparisons with other instruments. The observations provide vacuum-ultraviolet (VUV) radiometry of the Sun in many emission lines and continua of atoms and ions with relative standard uncertainties of 15 % (detector A) and 20 % (detector B) for the wavelength range 53 nm to 124 nm, with larger uncertainties outside this interval and after the SOHO recovery in We report on the present state of the SUMER radiometric calibration and provide a full bibliography related to this topic Introduction A quantitative investigation of the solar VUV radiation is of great interest because its variation is closely related to solar magnetic activity. The VUV radiation incident on 145

2 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER the Earth is completely absorbed by the high-altitude atmosphere. Consequently, it controls many processes in the upper atmosphere; this is the reason why knowledge of the irradiance in this wavelength range is of importance. Shortward of nm, the O 2 photoionization limit, the VUV is responsible for the daytime ionosphere and its heating. This absorption also leads to other important processes in the Earth s ionosphere, including photoionization of N 2, O and NO at wavelengths shorter than (79.6, 91.1, and 134.0) nm, respectively, as well as photodissociation of N 2 shortward of nm and O 2 shortward of nm. The solar VUV spectrum is highly variable on all time scales. It varies, for instance, with the Sun s rotation period and the solar sunspot cycle. The variability during a solar rotation is to a large extent controlled by the contributions of active regions. The emission lines in this wavelength range are formed at electron temperatures between K and several million kelvin allowing us to probe the solar atmosphere from the chromosphere through the transition region to the corona. The continuum radiation near 150 nm is emitted close to the temperature minimum at the top of the photosphere. The aim of this contribution is to present the radiometric status of SUMER on SOHO after more than five years of successful solar observations. SOHO [Domingo et al., 1995] was launched on 2 December 1995 and has been observing the Sun from the inner Sun- Earth Lagrange point (L1, about km sunwards from the Earth) since early During the first years of operation, the solar activity was near its minimum between sunspot cycles 22 and 23 and reached the next maximum in The observations were interrupted from June until October A loss of the spacecraft attitude led to totally-uncontrolled thermal conditions. The SUMER subsystems, for instance, experienced severe temperature excursions from the design values: in the case of the telescope mirror a decrease of more than 100 C has been estimated for this radiometrically-critical item Instrument Description SUMER and its calibration concept have been described elsewhere [Schühle, 1994; Wilhelm et al., 1995], and its performance characteristics, after several months of spacebased operation, have been summarized by Wilhelm et al. [1997a] and Lemaire et al. [1997]. Both SUMER detectors A and B have photocathode areas with different responsivities: the central part of its microchannel plate (MCP) has a coating of potassium bromide (KBr) and the other parts use the bare plate. The angular pixel size in the direction along the slit is approximately 1 ( rad), corresponding to about 715 km on the Sun. The spectral resolution element in first order is about 4.4 pm, and 2.2 pm in second order. The instrument can perform spectral scans in the wavelength range 66 nm to 161 nm (in first order) and from 46.5 nm to 80.5 nm (in second order: the lower limit is determined by the low reflectivity of the normal-incidence optics at short wavelengths). The first-order and the second-order spectra are superimposed (some lines are seen even in third order) and can be recorded at pointing positions selected on the disk or above the limb in the low corona. With the slits oriented in the north-south direction, raster images of the full solar disk (or parts of it) and of the inner corona have been obtained by scanning the telescope mirror. Scans were performed in angular step sizes of multiples of across the disk and in eight swaths for most of the full-sun images.

3 10.3. Radiometric Calibration Radiometric Calibration The radiometric calibration and the long-term responsivity tracking of an instrument designed to take measurements on spacecraft pose specific difficulties. The laboratory calibration of SUMER was performed with the help of a transfer source standard [Hollandt et al., 1992, 1993, 1996a, b] traceable to BESSY I as a primary source standard. It was completed in early Extreme particulate and chemical cleanliness requirements [Schühle, 1993] had to be fulfilled during this operation and during the spacecraft integration activities [Thomas, 2002]. An instrument door and an electrostatic solar-wind deflector in front of the primary telescope mirror are specific features incorporated in the design in order to maintain the radiometric responsivity during the launch and operational phases. First VUV observations of the Sun were performed with SUMER in January Initial Characterization The laboratory calibration was primarily intended to measure the radiometric response of the system, as far as mirror reflectivities and detector responsivities were concerned. This was achieved without internal vignetting of the incident calibrated beam. An assessment of the relative standard uncertainties of the SUMER system and its subsystems for measurements in the spectral range from 53.7 nm to nm, which could be reliably calibrated in the laboratory, has been presented in table 1 of Wilhelm et al. [2000a]. It resulted in relative uncertainties of 11 % using a 2 mm hole in place of the slit, and 12 % with slit # 2 (1 300 ), where some vignetting by the slit had to be accepted in order to increase the spectral resolution and to reduce the detector counting rates [Hollandt et al., 1996a]. We find a relative uncertainty of approximately 18 % for the smallest slit ( ). When applying the spectral responsivity curves shown in Figure 10.1(b), (c), (d), which are based on the most recent assessment of the performance of the instrument (discussed below), to solar measurements, it is therefore necessary to take into account the vignetting effects of field stops as well as the diffraction at the slit. This can be accomplished by applying the SUMER calibration programme radiometry.pro. This programme can be found in the software tree of Solarsoft at soho/sumer/idl/contrib/wilhelm/rad/. It is under configuration control. Although the responsivities and, later on, the irradiance measurements are shown here in photon units, the programme is also designed to perform all calculations in energy units in accordance with the International System of Units (SI) [BIPM, 1998]. Note that the uncertainties given in panels (c) and (d) include the contributions of optical stops and diffraction effects. They can be considered as system characteristics of the SUMER configuration, but the uncertainty of the pixel size has to be treated separately. The relative uncertainty for detector B in the central wavelength range is larger, because this device is, in the interest of longevity, operated at a lower gain than during the ground calibration [Schühle et al., 2000a]. This reduced the responsivity by a factor of As detector B had a relative responsivity approximately 20 % higher than detector A during the laboratory calibration (cf., figure 7 of Wilhelm et al. [1995]), both detectors were expected to have the same responsivity in flight. In order to obtain a radiometric characterization in the spectral ranges at shorter and longer wavelengths, we compared line-radiance ratios measured on the solar disk with the results of atomic physics calculations. We also evaluated SUMER stellar observations and

4 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER Figure 10.1: Spectral responsivities of the SUMER instrument with its detectors A and B, and the corresponding relative uncertainties for the slit. (a) In the upper panel the responsivity ratio of the photocathodes is shown. (b) First-order and second-order KBr responsivities evaluated jointly for detectors A and B. The long-wavelength deviation of detector B is treated in the text as well as the bare MCP responsivities and the third-order calibration of detector A. The bars in the lower portion of this panel indicate the order of diffraction. (c) Independent detector A assessment, and (d) corresponding results for detector B. The relative uncertainties in the central wavelength range (indicated by vertical bars at 53 nm and 124 nm) are derived from the laboratory calibration and are smaller than those of the in-flight calibration extensions. The changes after the attitude loss of SOHO are indicated for both detectors. On the long-wavelength side, the relative uncertainty refers to the KBr photocathode only. The Interim scale of detector B is explained in the text.

5 10.3. Radiometric Calibration 149 obtained radiometric information with the help of the International Ultraviolet Explorer (IUE) and Hubble Space Telescope standard stars [Bohlin et al., 1990; Nichols and Linsky, 1996; Wilhelm et al., 1997b; Hollandt et al., 1998; Schühle et al., 2000a]. The responsivity ratios of the KBr and bare photocathodes in panel (a) have been measured in the laboratory and, in a wider wavelength range, during the mission operations. A very consistent and stable spectral dependence has been found employing various methods [Wilhelm et al., 2000a]. In the past, both detectors were treated independently. This type of evaluation is continued here, and the latest results are shown in panels (c) and (d). The main change compared to earlier evaluations is related to the post-loss responsivity (cf., Section ). Up to now responsivity functions were only available in first and second order. A deepexposure reference spectrum obtained with detector A in a stable coronal streamer on 13 and 14 June 2000 showed the Si XII pair at nm and nm in second and third order of diffraction. This allows us to establish responsivity curves in third order for both photocathodes. As the technical parameters of SUMER restrict observations of spectral lines in third order to the narrow range of 46.5 nm to 53.6 nm, the Si XII results were, with some confidence, extrapolated over this range as shown in Figure 10.1(b) Calibration Tracking Unless a space-borne instrument can be recovered after flight and re-calibrated, its calibration stability is a major concern during the launch phase and the mission duration. Comparison with other instruments offers a solution that we have used extensively. In addition to the IUE data mentioned above, which agreed with SUMER observations near 120 nm, where both instruments overlap in their wavelength ranges, comparisons with the Coronal Diagnostic Spectrometer (CDS) on SOHO [Harrison et al., 1995] and with the SOLar-STellar Irradiance Comparison Experiment (SOLSTICE) on the Upper Atmosphere Research Satellite (UARS) [Rottman et al., 1993] have been considered in this context. The CDS indicated higher radiance values of quiet-sun regions than SUMER, but the relative deviations of about 30 % for He I at 58.4 nm and 9 % to 17 % for Mg X at (60.9 and 62.5) nm are within the combined uncertainty margins [Pauluhn et al., 1999]. An important aspect is that the ratio of the radiances derived from both instruments remained rather constant over several months, although relative variations in radiance of ±20 % occurred. Consequently, we can conclude that the variations are of solar origin [Pauluhn et al., 2001a, b]. SOLSTICE solar irradiance measurements in N V at nm resulted in values a factor of 1.14 higher than those obtained from SUMER full-disk rasters and only by 1.11 higher in C IV at nm [Wilhelm et al., 1999a]. Given the good radiometric calibration and stability of SOLSTICE this seemed to indicate that the SUMER response was too small by a factor just over 1.1 on the long-wavelength side at that time. In a comparison between a coronal spectrum obtained by SUMER and the CHIANTI database, Landi et al. [2002] found that the observed ratio of the Si XII lines at 49.9 nm and 52.0 nm agrees very well with the theoretical predictions. Moreover, there is no evidence for any other gross inaccuracy in the full SUMER wavelength range. Doschek et al. [1999] also compared theoretically-calculated line ratios of several ions with SUMER observations. In many cases, excellent agreement was found, but also significant discrepancies, even for ratios in narrow wavelength intervals. Consequently, the main conclusion

6 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER was that improvements in atomic data calculations are required, in particular, for sulfur ions. A stable radiometric calibration is also supported by the results of the SUMER flatfield exposures performed in the H I Lyman continuum near 88 nm. Quiet-Sun conditions have been selected by pointing to appropriate areas. No significant change of the responsivity of detector A has been found over 200 d, nor was there any decrease in detector B over 350 d, but a relative increase of less than 20 % could not be ruled out [Schühle et al., 1998]. In addition, the stable flat-field counting rates, in particular in the case of detector A, indicate that the radiometric characterization could be maintained, although the high voltage applied for the operation of the MCPs had to be raised by more than 500 V to achieve a constant gain. However, the recent detection of two slight inconsistencies has led to an update of the SUMER radiometric calibration in two areas: (1) a comparison of the full-sun data of detector A with those of detector B in S VI at 93.3 nm and H I at 93.7 nm showed that a test period had to be introduced in radiometry.pro at the end of September 1996 (see Section 10.4); (2) stellar observations [see Lemaire, 2002] and recent reference spectra of quiet-sun regions taken on both detectors indicate that the responsivity of detector B is higher than detector A longward of 104 nm. However, the slope of the increase is different, in that the stellar observations show a factor of nearly two at 147 nm, whereas the reference spectra give only 1.31 at 150 nm. Work is ongoing to resolve this discrepancy. At this stage we provide both options in the SUMER radiometry evaluation. Firstly, we assume the reference spectrum result and re-scale detector B accordingly. It then transpires that the data points available for both detectors are not systematically different within the relative uncertainty margin of 20 %. It has thus been possible to determine joint KBr responsivity functions for detectors A and B in Figure 10.1(b). The detector B excess longward of 104 nm has, of course, to be taken into account again later. Finally, with the help of the KBr-to-bare ratio, responsivities can be obtained for the bare MCP. Secondly, we adopt the ratio obtained with the star α Leo which was the basis for the previous, separatelyevaluated responsivities of both detectors. See Section 10.5 for the actual implementation in radiometry.pro Other Aspects Relating to Radiometry The radiometric characterization of SUMER has been performed in such a way that averages over many detector pixels have been used to determine the responsivity [Wilhelm et al., 1995; Hollandt et al., 1996a]. Because of relative variations of the pixel size (2 % in the spectral dimension and 1.5 % in the spatial dimension), and a varying flat-field response [Dammasch et al., 1999a; Schühle et al., 2000a], the single-pixel data have to be treated with great care. The spectral responsivity curves displayed in Figure 10.1(b), (c), (d) refer to situations with low counting rates both for the total detector and for single pixels. Whenever the total counting rate exceeds about s 1, a deadtime correction is required and with a single-pixel rate above about 3 s 1, a gain-depression correction is called for (cf., Section 10.5). The responsivity curves are the best functions available to date for the period from January 1996 to June 1998 and from November 1998 to December 2001, but they have undergone modifications in the past and might do so again in the future. There are two

7 10.3. Radiometric Calibration 151 Figure 10.2: Continuation of the history of the SUMER radiometric calibration revisions during the SOHO mission until 7 February The period from December 2001 to 7 February 2002 (Epoch 8) is not shown as it deviates from the recent one only by a slight modification of the interpolation method between 70 nm and 80 nm on the bare MCP. The calibration status before October 1999 is documented in Figure 4 of Wilhelm et al. [2000a]. The ratios (shown on a scale between 0.33 and 3) of the responsivity functions to the most recent functions are shown for both detector A (KBr: solid lines; bare: dotted lines) and detector B (KBr: dashed lines; bare: dashed-dotted lines) in both orders of diffraction. The epochs applied are given in each panel. The best calibrations available now are represented by unity in panel (d) for the separate evaluation in panels (a) to (c) and in (h) for the joint one in panels (e) to (g). In panels (d) and (h), the reduction of responsivity caused by the attitude loss is also indicated. The test phase of detector B in 1996 could not be documented in the format of this diagram.

8 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER reasons for such modifications: an improved understanding of the performance of the instrument including its calibration, and changes of the status of the instrument with time. Figure 10.2 summarizes the modifications since October 1999 continuing the history documented in figure 4 of Wilhelm et al. [2000a]. The plots demonstrate that the calibration status in the central wavelength range is very consistent over time for both detectors. The changes apparent at long wavelengths in panels (e) to (g) for detector B are related to the inconsistencies of the star data mentioned above. We recommend the use of the joint evaluation in this range. Most of the other small changes present stem from modifications of the interpolation method. The bulge near 80 nm for detector B seems to indicate that in this interval both detectors cannot be treated jointly. It might be advisable to use the separate responsivity of detector B here. At short wavelengths, the deep exposure in the corona furnished, in addition to the third-order measurements of the Si XII line pair, precise second-order data that led to a modification for detector A. The deep exposure also proved that the Ne VII line at nm in second order cannot be separated from H I nm and H II nm in first order. It was, therefore, eliminated from the data set. Before an adequate low-gain level of detector B was found, a test configuration was used between 24 September and 6 October 1996, which will be considered below. The revisions shown in Figure 10.2(a) to (c) and (e) to (g) resulted from a better understanding of the calibration status. However, a major change of the responsivity happened during the attitude loss of SOHO. After the recovery of SOHO, a responsivity decrease was found over a wide spectral range as shown in panels (d) and (h) by the dashed curves. We attribute this change to the deposition of contaminants and subsequent polymerization on the optical surfaces of the instrument, because both detectors were equally affected. The telescope mirror is considered to be the prime candidate. This critical item was held at an elevated temperature of about 40 C (non-operational) or about 80 C (operational) compared with 23 C of the optical bench structure at all times of the nominal mission. This was done in order to avoid any condensation on the mirror. During the interval of SOHO s uncontrolled attitude, however, the temperature estimate for this mirror was near 80 C, whereas other parts of the spacecraft reached very high temperatures. A first assessment led to a decrease in relative responsivity of 43 % [Wilhelm et al., 2000a]. Schühle et al. [2000b], in a second study, included only observations which could not have been influenced by active-region contributions. Relative losses of 26 % for He I 58.4 nm, 28 % for Mg X (60.9 and 62.4) nm, 34 % for Ne VIII 77.0 nm, 39 % for N V nm, and 29 % for the H I Lyman continuum were obtained, resulting in an average relative loss of 31 %. It should be noted that star observations of α Leo before and after the attitude loss provide strong evidence that the responsivity decrease is wavelength dependent with a tendency to become rather small at the longest wavelengths. Additional observations were performed in August 2001 [Lemaire, 2002]. Reference spectra taken in quiet-sun regions furnish yet another means of determining the wavelength dependence of instrumental changes. The small solar areas covered by the slit during the observation of reference spectra and the relatively short integration times lead, together with significant variations of the solar radiance even of the quiet-sun chromospheric network, to considerable scatter and uncertainties. However, these observations confirm that the longer wavelengths of the SUMER range are less affected by the attitude loss. Consequently, we adopted a relative loss in the responsivities of 31 % for wavelengths shorter than nm (N V) (as before), which linearly decreases to 5 % at 161 nm.

9 10.4. Radiances and Irradiances 153 Quiet-Sun reference spectra obtained with both detectors also provide consistency checks on the responsivity functions of detectors A and B, which have been established independently, as well as on the KBr-to-bare photocathode ratio. No significant change of this ratio could be detected over the operational period since the laboratory measurements (cf., Figure 10.1(a)) Radiances and Irradiances A first-order solar radiance spectrum obtained on 12 August 1996 and treated with the most recent calibration software is displayed in Figure In addition to the emission lines, several continua are very prominent. Some black-body radiation levels have been plotted by adjusting the temperature in Planck s law. The temperature required to produce an adequate fit at long wavelengths is close to the values expected near the temperature minimum in the solar atmosphere [cf., Brekke and Kjeldseth-Moe, 1994; Samain et al., 1975], thus supporting the responsivity curves applied. By studying the line radiance measurements of quiet-sun areas performed by the Joint Observing Programme (JOP) named ICAL 01, which was designed for the inter-calibration of SOHO instruments, a significant increase in all emission lines has been found from 1996 onwards [Schühle et al., 2000b, c, d; Wilhelm et al., 2000a]. In evaluating spectra of the type shown in Figure 10.3, the overlapping first and second orders of diffraction cause some difficulties. For emission lines the separation is relatively straightforward (provided their orders have been established). The second-order lines have in fact been manually removed from the spectrum in Figure However for continua the task is not so trivial. In principle, observations on both photocathodes could be used to separate the diffraction orders, but at long wavelengths, where the second-order contribution is most significant, the KBr-to-bare ratios are close to one in both orders (cf., Figure 10.1(a)), and the results are rather uncertain. A complete quiet-sun reference spectrum observed on detector A (78 to 160) nm on 12 November 2001 and another one on detector B (67 to 149) nm shortly afterwards, allowed us to separate the continua between 134 nm and 160 nm, where second-order contributions are of importance. The first-tosecond-order responsivity ratio of the grating had been found to be 1.66 at 79.0 nm (O IV) for detector A [Wilhelm et al., 1997a]. This ratio could be confirmed with the recent B spectrum for this line, but the lines Ne VIII 77.0 nm and N V 76.5 nm indicate a decrease of the ratio with decreasing wavelength. At 70.4 nm (O III), we find Under the assumption of a linear decrease of the ratio with wavelength, we can then use the spectra obtained to find the contribution of the second-order continuum to the first-order observations at long wavelengths. The results are plotted in Figure 10.4 for both photocathodes. Solar irradiances of emission lines can be derived from SUMER measurements without assumptions only when a raster scan of the whole disk has been performed. This has been done for many lines. Since line irradiances derived from full-sun rasters provide the best opportunities for radiometric inter-comparisons, we present in Table 10.1 a selection of such observations treated with the latest SUMER calibration routine (epoch 9), normalized to 1 AU (astronomical unit). Typical relative standard uncertainties are less than 20 % for most of the emission lines, but 30 % for C IV. Note that the agreement with SOLSTICE in Section was much better than these wide margins, indicating that the SUMER uncertainties, in particular at long wavelengths, are probably too conservative.

10 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER Figure 10.3: SUMER radiance spectrum of a quiet-sun region in first order of diffraction (second-order spectral lines have been removed manually). Prominent lines and continua are identified. Some black-body radiation levels are shown to provide an estimate of the radiation temperatures of the continua. A SUMER spectral resolution element in first order corresponds to 7 µm on the scale of this diagram. It should be mentioned that the latest calibration yields a relative SOLSTICE/SUMER agreement for C IV at nm within 4 %. All full-disk images were taken with the slit # 2 (1 300 ) and most of them on detector A. These images allow us to determine the limb brightening of the corresponding emission lines, and also to separate the irradiance into contributions from quiet-sun areas, coronal holes or active regions [Wilhelm et al., 1998]. In the case of C III, missing parts of the disk have been restored with average disk values up to the photospheric limb. In previously calculated and published SUMER values [Wilhelm et al., 1998, 1999b, 2000b] some of the line radiances seem to have been underestimated. This could be due to an attempt to obtain a pure image in a spectral line. New, detailed studies of the profiles led to the results listed in Table For this refinement of the evaluation, gain and deadtime corrections were applied as indicated. For the He I line, a relative correction of about 9 % is now performed based on the actual detector counting rates. This is different from the earlier evaluations mentioned and explains the higher value obtained. The Ne VIII line has

11 10.4. Radiances and Irradiances 155 Figure 10.4: Fractional contribution of the H I Lyman continuum to the sum of the firstand second-order continuum at long wavelengths as observed on both photocathodes in reference spectra obtained with both detectors on 12 November been observed in second order of diffraction, where it is hopelessly blended with weak Si I lines in first order, whose relative contributions, however, amount only to a few percent. With the exception of O V, which was integrated on board, all other results are well within the declared uncertainty margins of the earlier results. We have thus studied the O V 63.0 nm line in greater detail using radiance data from inter-calibration runs, reference spectra and other quiet-sun observations in 1996 together with measured centre-to-limb variations (limb-brightening factor of 3.7), and find an estimate for the average line irradiance in photons of s 1 m 2 with a relative standard uncertainty of 20 %. This value agrees with irradiances obtained from the full-sun data within the uncertainty margins. An average of s 1 m 2 was obtained from observations in the years 1997 to There was no systematic difference between observations on the KBr photocathode and the bare MCP indicating that the second order O V line is not blended by any prominent first-order line. During a 360 roll manœuvre of SOHO on 20 March 1997, the SUMER slit was set to cross the solar limb and then moved around the disk with the spacecraft rotation. Together with coordinated quiet-sun radiance measurements, irradiance values for many emission lines could be estimated with a minimum of assumptions [Dammasch et al., 1999b; Schühle, 2000c]. Taking the N V line at nm as an example, the result in photons was E(N V) = s 1 m 2 with a relative standard uncertainty of 25 %. An overview of irradiance measurements and estimates obtained with SUMER is shown in Figure Also included is an O V photon irradiance estimate at s 1 m 2 obtained during the same roll manœuvre. This value is more than a factor of 1.3 higher than the full-sun irradiance in Inspection of the corresponding centre-to-limb variations showed significant enhancements at or near the limb in 1997 caused by small active regions. In assessing the consistency with other O V irradiance measurements, it must be taken into account that the method of estimating the irradiance from the roll data introduces considerable additional uncertainties. Estimates for thirty-three other emission lines are available from this roll as well.

12 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER Table 10.1: Observing parameters and irradiance results of complete solar images in the C III, Ne VIII, C IV, and He I emission lines on detector A with high spatial and the full SUMER spectral resolution. The O V window was spectrally integrated (i) over 50 pixels on board of SOHO, but corrected for Si I blends and continuum contributions (combined relative correction: 37 %). Spectral line C III Ne VIII C IV He I O V Wavelength, λ/nm a a 63.0 a Temperature, T e /kk Exposure time, t s /s Angular step width, s/ Photocathode bare bare KBr bare KBr Spectral window in pixel (i) Date in Jan 02 Feb 04 Feb 02 Mar 14 Jun Solar radius from Earth b, R / Start time in UTC 22:40 02:05 02:06 23:22 11:46 Duration in h:min 25:27 16:03 31:34 27:24 3:44 F 10.7 /(10 22 W m 2 Hz 1 ) Counting rate c, N c /(10 6 s 1 ) Line irradiance at 1 AU E/(µW m 2 ) in photons, E/(10 12 s 1 m 2 ) Relative gain correction, c g /% Relative deadtime correction, c d /% Relative standard uncertainty, u r /% a Observed in second order of diffraction. b Seen from SOHO the radius is larger by a factor of c If the full Sun were to be seen by the detector. Another roll manœuvre was performed on 14 and 15 November SUMER was operated in a mode similar to that in Quick-look data confirmed that good observations have been obtained, but they could not be evaluated in time for this report. Without considering the test period between 24 September and 6 October 1996 mentioned above and the resulting additional correction, the S VI and H I irradiances would be too low by a factor of 1.18, which would introduce a discontinuity much larger than the variations before that date suggest. A detailed study of the dependence of the responsivity of detector B on the gain settings (level of the high voltage) showed that half of this correction can be understood as a consequence of the specific test configuration. The remaining portion of the reduction could not be explained, but is well within the uncertainty margin anyway. The correction can be applied by setting the keywords test and det b in radiometry.pro. It is shown in Figure 10.1(d) by the Interim scale. Any evaluation of observations in the above time interval performed without the keyword test has to be corrected accordingly.

13 10.5. Conclusions 157 Figure 10.5: Solar irradiance observations in VUV emission lines with SUMER in 1996 and Detector B was used after 23 September Not all uncertainty margins are shown in the interest of clarity. Spectral lines with wavelengths shortward of 79 nm have been observed in second order Conclusions SUMER data can be obtained for further investigations from the public domain of the SOHO archive ( html) or the SUMER Image Database ( projekte/sumer/file/sumerentrypage.html). Both SOHO and SUMER are still operational, albeit the latter with reduced performance capabilities, in particular concerning spatial rasters and the remaining count potential of the detectors. The responsivity curves in Figure 10.1 present the status before and after the SOHO attitude loss. The radiometric calibration of SUMER data requires the application of the programme radiometry.pro, of which the latest and official version is that contained in the software tree of Solarsoft at any given time. The December 2001 and February 2002 issues (epoch 8 and epoch 9) contain two slightly-different versions according to their evaluation (separate or joint) except for the bare MCP in second order. We recommend the use of the joint alternative, when available, as it provides the best consistency between detectors A and B. However, in the interest of a complete change history of the independently-calibrated detectors the sequence of responsivity curves is also

14 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER continued in the old format, and, as mentioned in Section , the separate evaluation should be used for detector B between 75 nm and 88 nm. Any modifications have been and will be announced to the user community, and are also documented in the comments accompanying the programme. An update of the history of the responsivity functions between October 1999 and December 2001 has been presented in Figure With the help of epoch keywords, all past calibration stages can be recovered. Other keywords control the specific responsivity function to be applied. Programmes for deadtime corrections (deadtime corr.pro) and the reversal of local gain depressions (local gain corr.pro) are also available for observations with high counting rates. Given the calibration activities with SUMER outlined above, we have achieved a relative standard uncertainty of 15 % to 20 % in the wavelength range from 53 nm to 124 nm for high-resolution measurements in the spatial and spectral regimes. This is better than we could have anticipated in the planning phase of SOHO and SUMER. This success of the SUMER calibration, and of SOHO in general, will also be an important benchmark for future missions. Acknowledgements SUMER is financially supported by DLR, CNES, NASA and the ESA PRODEX programme (Swiss contribution). We thank C.D. Pike and the referee, R.A. Harrison, for many valuable comments and suggestions. Bibliography BIPM (Bureau International des Poids et Mesures), Le Système International d Unités (SI) and The International System of Units, 7th ed., Sèvres, France, Bohlin, R.C., Harris, A.W., Holm, A.V., and Gry, C., The ultraviolet calibration of the Hubble Space Telescope. IV - Absolute IUE fluxes of Hubble Space Telescope standard stars, Astrophys. J. Suppl. Ser. 73, , Brekke, P., and Kjeldseth-Moe, O., New radiometric values of the solar ultraviolet continuum radiation from 1500 to 1700 Å, Astrophys. J. 413, L55 L58, Dammasch, I.E., Wilhelm, K., Curdt, W., and Hassler, D.M., The Ne VIII (λ770) resonance line: Solar wavelengths determined by SUMER on SOHO, Astron. Astrophys. 346, , 1999a. Dammasch, I.E., Wilhelm, K., Curdt, W., and Schühle, U., Solar ultraviolet irradiance and radiance observations by SUMER on SOHO, ESA SP-448, , 1999b. Domingo, V., Fleck, B., and Poland, A.I., The SOHO mission: An overview, Sol. Phys. 162, 1 37, Doschek, E.E., Laming, J.M., Doschek, G.A., Feldman, U., and Wilhelm, K., A comparison of measurements of solar extreme-ultraviolet spectral line intensities emitted by C, N, O, and S ions with theoretical calculations, Astrophys. J. 518, , Harrison, R.A., Sawyer, E.C., Carter, M.K., Cruise, A.M., Cutler, R.M., Fludra, A., Hayes, R.W., Kent, B.J., Lang, J., Parker, D.J., Payne, J., Pike, C.D., Peskett, S.C., Richards, A.G., Culhane, J.L., Norman, K., Breeveld, A.A., Breeveld, E.R., Al Janabi, K.F., Mc- Calden, A.J., Parkinson, J.H., Self, D.G., Thomas, P.D., Poland, A.I., Thomas, R.J.,

15 Bibliography 159 Thompson, W.T., Kjeldseth-Moe, O., Brekke, P., Karud, J., Maltby, P., Aschenbach, B., Bräuninger, H., Kühne, M., Hollandt, J., Siegmund, O.H.W., Huber, M.C.E., Gabriel, A.H., Mason, H.E., and Bromage, B.J.I., The Coronal Diagnostic Spectrometer for the Solar and Heliospheric Observatory, Sol. Phys. 162, , Hollandt, J., Kühne, M., and Huber, M.C.E., Radiometric calibration of solar space telescopes - The development of a vacuum-ultraviolet transfer source standard, ESA Bulletin 69, Hollandt, J., Huber, M.C.E., and Kühne, M., Hollow cathode transfer standards for the radiometric calibration of VUV telescopes of the Solar and Heliospheric Observatory (SOHO), Metrologia 30, , Hollandt, J., Schühle, U., Paustian, W., Curdt, W., Kühne, M., Wende, B., and Wilhelm, K., Radiometric calibration of the telescope and ultraviolet spectrometer SUMER on SOHO, Appl. Opt. 35, , 1996a. Hollandt, J., Kühne, M., Huber, M.C.E., and Wende, B., Source standards for the radiometric calibration of astronomical telescopes in the VUV spectral range, Astron. Astrophys. Suppl. Ser. 115, , 1996b. Hollandt, J., Schühle, U., Curdt, W., Dammasch, I.E., Lemaire, P., and Wilhelm, K., Solar radiometry with the telescope and vacuum-ultraviolet spectrometer SUMER on the Solar and Heliospheric Observatory (SOHO), Metrologia 35, , Landi, E., Feldman, U., and Dere, K.P., CHIANTI An atomic database for emission lines. Paper V: Comparison with an isothermal spectrum observed with SUMER, Astrophys. J. Suppl., in press, Lemaire, P., SUMER stellar observations to monitor responsivity variations, this volume, Lemaire, P., Wilhelm, K., Curdt, W., Schühle, U., Marsch, E., Poland, A.I., Jordan, S.D., Thomas, R.J., Hassler, D.M., Vial, J.-C., Kühne, M., Huber, M.C.E., Siegmund, O.H.W., Gabriel, A., Timothy, J.G., and Grewing, M., First results of the SUMER telescope and spectrometer on SOHO. II. Imagery and data management, Sol. Phys. 170, , Nichols, J.S., and Linsky, J.L., The final archive and recalibration of the International Ultraviolet Explorer (IUE) satellite, Astron. J. 111, , Pauluhn, A., Rüedi, I., Solanki, S.K., Lang, J., Pike, C.D., Schühle, U., Thompson, W.T., Hollandt, J., and Huber, M.C.E., Intercalibration of SUMER and CDS on SOHO. I. SUMER detector A and CDS NIS, Appl. Opt. 38, , Pauluhn, A., Solanki, S.K., Schühle, U., Wilhelm, K., Lang, J., Thompson, W.T., Rüedi, I., Hollandt, J., and Huber, M.C.E., Comparison of quiet-sun radiances measured by SUMER and CDS on SOHO, Space Sci. Rev. 97, 63 66, 2001a. Pauluhn, A., Rüedi, I., Solanki, S.K., Schühle, U., Wilhelm, K., Lang, J., Thompson, W.T., Hollandt, J., and Huber, M.C.E., Intercalibration of SUMER and CDS on SOHO. II. SUMER detectors A and B and CDS NIS, Appl. Opt. 40, , 2001b. Rottman, G.J., Woods, T.N., and Sparn, T.P., Solar-Stellar Irradiance Comparison Experiment Instrument design and operation, J. Geophys. Res. 98, , Samain, D., Bonnet, R.M., Gayet, R., and Lizambert, C., Stigmatic spectra of the Sun between 1200 Å and 2100 Å, Astron. Astrophys. 39, 71 81, Schühle, U., The cleanliness control program for SUMER/SOHO, In: UV and X-Ray Spectroscopy of Astrophysical and Laboratory Plasmas, Proc. 10th International Colloquium, (eds. E.H. Silver and S.M. Kahn), Cambridge University Press, , 1993.

16 SOLAR VACUUM-ULTRAVIOLET RADIOMETRY WITH SUMER Schühle, U., The SUMER instrument on SOHO: Design, performance predictions, and calibration aspects, Proc. SPIE 2283, 47 52, Schühle, U., Brekke, P., Curdt, W., Hollandt, J., Lemaire, P., and Wilhelm, K., Radiometric calibration tracking of the vacuum-ultraviolet spectrometer SUMER during the first year of the SOHO mission, Appl. Opt. 37, , Schühle, U., Curdt, W., Hollandt, J., Feldman, U., Lemaire, P., and Wilhelm, K., Radiometric calibration of the vacuum-ultraviolet spectrograph SUMER on the SOHO spacecraft with the B detector, Appl. Opt. 39, , 2000a. Schühle, U., Hollandt, J., Pauluhn, A., and Wilhelm, K., Mid-term radiance variation of far-ultraviolet emission lines from quiet-sun areas, ESA SP-463, , 2000b. Schühle, U., Wilhelm, K., Hollandt, J., Lemaire, P., and Pauluhn, A., Radiance variations of the quiet Sun at far-ultraviolet wavelengths, Astron. Astrophys. 354, L71 L74, 2000c. Schühle, U., Pauluhn, A., Hollandt, J., Lemaire, P., and Wilhelm, K., Radiance variations of vacuum-ultraviolet emission lines of the quiet Sun observed with SUMER on SOHO, Phys. Chem. Earth (C) 25, , 2000d. Thomas, R., 20:20 vision; SOHO cleanliness, this volume, Wilhelm, K., Curdt, W., Marsch, E., Schühle, U., Lemaire, P., Gabriel, A.H., Vial, J.- C., Grewing, M., Huber, M.C.E., Jordan, S.D., Poland, A.I., Thomas, R.J., Kühne, M., Timothy, J.G., Hassler, D.M., and Siegmund, O.H.W., SUMER Solar Ultraviolet Measurements of Emitted Radiation, Sol. Phys. 162, , Wilhelm, K., Lemaire, P., Curdt, W., Schühle, U., Marsch, E., Poland, A.I., Jordan, S.D., Thomas, R.J., Hassler, D.M., Huber, M.C.E., Vial, J.-C., Kühne, M., Siegmund, O.H.W., Gabriel, A., Timothy, J.G., Grewing, M., Feldman, U., Hollandt, J., and Brekke, P., First results of the SUMER telescope and spectrometer Solar Ultraviolet Measurements of Emitted Radiation on SOHO. I. Spectra and spectroradiometry, Sol. Phys. 170, , 1997a. Wilhelm, K., Lemaire, P., Feldman, U., Hollandt, J., Schühle, U., and Curdt, W., Radiometric calibration of SUMER: Refinement of the laboratory results under operational conditions on SOHO, Appl. Opt. 36, , 1997b. Wilhelm, K., Lemaire, P., Dammasch, I.E., Hollandt, J., Schühle, U., Curdt, W., Kucera, T., Hassler, D.M., and Huber, M.C.E., Solar irradiances and radiances of UV and EUV lines during the minimum of the sunspot activity in 1996, Astron. Astrophys. 334, , Wilhelm, K., Woods, T.N., Schühle, U., Curdt, W., Lemaire, P., and Rottman, G.J., The solar ultraviolet spectrum from 1200 Å to 1560 Å: A radiometric comparison between SUMER/SOHO and SOLSTICE/UARS, Astron. Astrophys. 352, , 1999a. Wilhelm, K., Lemaire, P., Dammasch, I.E., Hollandt, J., Schühle, U., Curdt, W., Kucera, T., Hassler, D.M., and Huber, M.C.E., Solar irradiances of UV and EUV lines during the minimum of the sunspot activity in 1996, Adv. Space Res. 24, , 1999b. Wilhelm, K., Schühle, U., Curdt, W., Dammasch, I.E., Hollandt, J., Lemaire, P., and Huber, M.C.E., Solar spectroradiometry with the telescope and spectrograph SUMER on the Solar and Heliospheric Observatory SOHO, Metrologia 37, , 2000a. Wilhelm, K., Lemaire, P., Dammasch, I.E., Hollandt, J., Schühle, U., Curdt, W., Kucera, T., Hassler, D.M., and Huber, M.C.E., Solar irradiances of ultraviolet emission lines measured during the minimum of sunspot activity in 1996 and 1997, Phys. Chem. Earth (C) 25, , 2000b.